Testing and Control of Cerebrospinal Fluid Shunts

Abstract

In hydrocephalic patients, increased cerebrospinal fluid (CSF) volume puts the brain tissue under abnormal strain, leading to a degeneration of brain tissue and function. Hydrocephalus is usually treated by the surgical placement of a CSF shunt, which aims at restoring normal CSF compliance by draining excessive fluid. The drainage rate of today’s CSF shunts is regulated by passive differential pressure valves. Thus, they work towards a constant pressure difference between cranium and peritoneum – the usual drainage site. As intracranial pressure (ICP) and intraperitoneal pressure are not normally linked, there are problems with over- and also underdrainage, especially related to postural changes. Therefore, there has long been the vision of a feedback-controlled active CSF shunt. However, we are yet to see such a smart shunt on the market, due to unproven feasibility, unclear physiology, and high regulatory burdens.
To reduce the cost and effort associated with testing new hardware and control concepts, a hardware-in-the-loop test bench was implemented that allows for fast, cost-effective, and repeatable testing of passive and active CSF shunts within an environment that is close to in vivo conditions. On the test bench, the shunt to be tested is placed in a dynamic in vitro setup that interfaces with a mathematical model of the relevant physiology. The model is evaluated numerically in real-time and accounts for posture dependent behavior and viscoelastic effects. With this, the test bench allows to investigate how shunts interact with CSF dynamics under the complex pressure environment seen due to postural changes.
To investigate how siphoning-induced overdrainage in upright posture can be avoided, CSF drainage rates, resulting CSF volume and ICP were determined for three different types of anti-siphon devices (ASD) in supine, sitting, and standing posture. This comparison study showed that today’s passive ASDs can prevent overdrainage. However, they cannot ensure continuous drainage independent of posture and intraperitoneal pressure.
Simulating the shunt’s in vivo environment on the test bench requires a precise mathematical model of the relevant physiology and how it changes with posture. To explain the changes in the craniospinal CSF volume and compliance distribution associated with posture changes, I implemented a lumped-parameter model of the CSF system and the relevant parts of the cardiovascular system. The good concordance of this model with clinical observations indicated that jugular collapse is a major contributor to CSF dynamic in upright posture.
Modelling how posture influences CSF dynamics has shown that the pulse pressure amplitude (AMP) of the ICP might be a suitable control variable for active CSF shunts, because it is theoretically independent of posture. However, calculating AMP from ICP recordings in supine and sitting posture has shown a significant AMP increase in sitting posture.
In conclusion, the implementation of a hardware-in-the-loop test bench can reduce the development costs of active CSF shunts by enabling realistic testing, without the need for in vivo experiments. With this, the test bench may catalyze the development of active shunts, which are currently prevented by their questionable economic viability. Comparing different shunts has already contributed to improving our understanding of the complex interaction of shunt and patient. Moreover, AMP, a potential control variable for active CSF shunts, was shown to be posture-dependent. But, the derivation of more sophisticated mathematical models has improved the understanding of how posture influences CSF dynamics.